Andrew: Hey, everyone. My name is Andrew Warner. I’m the founder of Mixergy, where I interview entrepreneurs about how they built their businesses.
Apart from this, the only other obsession that I have—well, two, I guess. Well, no, I have many obsessions, I guess. I’m obsessed with my family, and I’m starting this interview late because we had some family issues. My son was throwing up last night. It was really painful. So that takes up a lot of my time. Then I also try to get some workouts in, but I barely get a chance to do that. Then I do Mixergy interviews, and when I’m not doing this, the thing that’s really made me stressed lately is I’m taking on a lot of work around chatbots.
I’m so freaking obsessed with them. I know it’s eating away into my research time with my guests. I know it’s eating away at my family time. I know it’s keeping me from running because I think that this is the future of online engagement. I think just like we went from apps to web apps, and that was a dramatic change in the landscape. Then we went from desktop to mobile, and that was a dramatic change and new companies were born. I feel like chat as a platform is going to do the same thing, that we’re going to see new businesses born, new companies rise up because they know how to create experiences inside of chat apps.
Chat is being used more and more. Facebook Messenger alone is 1.3 billion people. WhatsApp has roughly the same number, and we’re spending more time in chat than we’re spending in just about any other medium, including social media, which sucks up a lot of our time. We’re spending more time chatting than we are in social media, more time in chat apps than we are in email. So businesses that know how to capitalize on that are going to win in the future, and those that don’t are going to watch the others take over. So I’ve been kind of obsessed with that.
Because you guys have heard me talk about this, you’re introducing me to companies that are doing really well in this. One of my old friends, Alexey Prohorenko, heard me get excited about this. He said, “You’ve got to meet my friend.” I said, “Who’s your friend?” He said, “My friend is a guy named Matthew Black. He’s had a couple of interesting companies, used to run this one business, Contextual, which he sold. He had this other business, CrowdHype, which he had a rough experience with.” But he said, “What he’s obsessed with, Andrew, like you is Matthew Black is obsessed with chatbots.”
I had our producer talk to him. This is a guy that had a business that was going well, a digital agency that was going well. He got so freaking obsessed with chatbots that he freaking lost it all. The revenue was going down. He went in debt and partially it’s because chatbots didn’t work out in the first iteration, the first way that we thought it was going to go.
But now, he’s back on top. It’s working. It’s actually getting him clients and it’s working for his clients. He’s building chatbots for business. I probed him before the interview started. Is it just that you’re selling it to them and they’re willing to pay you? He said, “No, they’re getting results.” He gave me some numbers. We’re going to talk about as many numbers as we can in this interview because chatbots are working for him and working for his clients.
I invited him here to talk about a few things. Number one, how does he get clients for this brand new thing that’s so hard to explain? Number two, how does he communicate to his clients’ users what a chatbot is. Number three, how is he getting results for his clients, and number four, where does he see the future of this medium that is so early stage that I think many of you are probably lost already just because I said the word “chatbot”?
All right. So we’re going to get into all of that. He is the founder of Black Ops. It’s a boutique agency that provides a suite of services focused on conversational marketing, and conversation is the word that we use now for chatbots because it’s more than just like chat in like Facebook Messenger. It’s going to evolve, and we don’t know where. It might be Alexa. It might be Google Home. It might be Siri. We don’t know exactly where, but conversational experiences are the future.
This interview is sponsored by two companies that I’ll tell you more about later. The first is Toptal, which will help you get your next great developer. The second will help you hire phenomenal designers. It’s called DesignCrowd. I’ll tell you more about both those later. Matthew, good to have you on here.
Matthew: Thank you, Andrew. What an intro, wow.
Andrew: I’m looking at my time as I go through the intro — four and a half minutes. I try to do two minutes. I’m like way over. You know what? Why don’t we jump to the end for a second here and give people a sense of size? Revenue-wise, where are you going to close 2017, since we’re three-fourths of the way into the year?
Matthew: Yeah. I think we’re going to close 2017 just under $1 million, and I think 2018 at least that, if not double that based on our pipeline.
Andrew: So just from building conversational experiences for your clients.
Matthew: Correct.
Andrew: At this point, it’s chatbots built on Facebook Messenger?
Matthew: Yeah. Facebook Messenger is probably 80% of what people come to us with. We definitely have seen a spike in SMS, which I can talk about a little bit later. Also, most recently Twitter Direct Message. They just opened up their platform, and the partners over at Twitter have been great to us. So we’re trying to educate our customers a little bit more on that as well.
Andrew: Let’s take a look at one of your clients, the debt company, what are they called?
Matthew: They’re called Debt Like WTF. Just to give you kind of a brief overview of them, they are a student loan debt consolidation service, which is a hot, hot market right now and they were around previously before we even talked. Essentially what they would do is they would use Facebook, they would target their ads towards people who they knew already went to college and they would rank colleges with the most debt. They would do demographics, 35 and under, and they would send them to their website, which was generally just a lead capture form, and then on the leads, they would smile and dial and try to walk them through their debt issue.
Andrew: They had their own employees?
Matthew: Yeah, two.
Andrew: Just two people on their team calling up and saying, “You’ve got trouble with your school debt, we will consolidate it, we will ease it up for you?”
Matthew: Two people, 12 hours a day, it was unreal. They were profitable. I think they were doing $20,000 to $30,000 a month these two guys just going bananas. It was funny too, because when they came to us, they thought the benefit of chatbots was simply to take the ease off these two guys. Operationally speaking, how can we make a chatbot make our process more efficient?
Andrew: What do you mean? Specifically what would the chatbot have done in their mind?
Matthew: So there’s two parts. There’s a qualifying to say this customer is actually a good customer for our service, and then there’s the fulfillment—collect their income, process the forms with the government, see exactly which loans are with which servicer and say, “Those are the ones we can help,” and then ultimately do the debt consolidation.
Andrew: They thought that before you get on a call with one of these two guys, you would talk to a chatbot, tell them how much debt you had, tell them how much trouble you were in for not paying it and maybe a couple of other questions, and if you had more than a certain amount of debt, if you qualified, then they would get on a call with you, that way the two guys weren’t wasting their time on people who weren’t qualified. Is that right? That’s what they—
Matthew: So that’s actually what we do now.
Andrew: What did they envision?
Matthew: They envisioned the chatbot servicing less as a qualifying tool and more as a fulfillment tool. What they realize is that given someone in that circumstance, I’m a 27-year old suffering with $30,000 worth of student debt, haven’t made a payment in four months, the last thing I want to be doing is hopping on a phone call and telling someone about it, right? So they had thought a chatbot, specifically one who lives on Facebook, would be a better way to service those customers. So preexisting customers, customers they already have, that would be the main type.
Andrew: I see. That chatbots are not ready to do well yet, am I right?
Matthew: It will really depend on who you ask, but I am in the camp that believes that servicing, operations and the majority of real customer service tasks, the artificial intelligence just isn’t there to provide a compelling experience.
Andrew: So you had to shift them away from that to today. How do they use the chatbot?
Matthew: Customer acquisition.
Andrew: So an ad, somebody sees the ad on Facebook. The ad is targeted to them. They have to be a certain age, probably have gone to school, etc. They see the ad, they click a button. Do they go directly to Facebook Messenger and start a conversation with the chatbot?
Matthew: Correct. We actually use some advanced functionality that does two-way. So the chatbot can actually talk to the ad, and the ad can actually talk to the chatbot. I know you had an episode on here previously where one of your guests went deep into structured JSON. So I probably won’t go that deep. But essentially we’re making really clever use of that kind of technology.
Andrew: That’s incredibly powerfully, right?
Matthew: Very.
Andrew: I remember when Dmitry of Chatfuel first told me about it, I didn’t understand it all. I’ve since played with it a little bit. It’s kind of cool. An ad can actually pop up a little chat window on people’s desktop computers. People can engage. All right. I’m kind of engaging in this Debt Like WTF chatbot right now. I just clicked the start on it. It says, “Hey there. I’m like the Debt Like . . .” I’m going to say the full word.
Matthew: Go for it.
Andrew: “I’m the Debt Like What the Fuck chatbot. I can help you save $$$ on your student loans.” “Okay.” “Let’s start with a few questions to help me understand how we can help you save some cash. Just click the quick reply button that pop up after each question.” So here’s the first question, “Do you have federal loans, Andrew?” This is there way of qualifying.
Matthew: Qualifying.
Andrew: I’m going to say, “I’m not sure.” If I said, “I don’t have federal loans,” I’m assuming I wouldn’t be a good fit. “Is your loan balance greater than $8,000?” I’m going to say I’m not sure because I don’t want to disqualify myself yet. “What is the status of your student loan? Is it in forbearance? Is it repayment?” Again, “I’m not sure.” It’s going to keep asking me all these questions. Here’s the next one, “Are you making payments to your student loans?” I’m going to say I’m not sure.
Okay. This is now qualifying and if I’m the right person, then I schedule a call with their people?
Matthew: Yeah. So it’s actually both. There’s two ways. So every question that you answer, I just want to point this out, it actually is a dynamic method to score you as a customer. So based on you saying, “I don’t know,” you probably have a low score on their end. It actually gets sent to their Slack real time. So it’s clever.
Andrew: There’s someone on the other side just watching their Slack and seeing that I’ve pressed these buttons.
Matthew: 100%.
Andrew: So if I start typing something in, they’ll get, I’m sure, an alert saying something outside of the usual buttons is happening. You might want to start a conversation with this guy, Andrew.
Matthew: Yeah. That all depends on the confidence score. The confidence score can either have the bot reply to you, pause the bot and allow one of the agents at their office to correct the flow of the bot, or simply just stop the bot, code read, go in there and manually take over.
Andrew: So if I were going to, instead of pressing these buttons, say, “I cannot sleep. I am in deep trouble right now. My wife and I are arguing over this. Is there someone who can help me?” They will see that, and chances are if they’re awake and watching, they could start talking to me right now and saying, “Calm down, Andrew, we’ve got you covered. Here’s our cellphone number, call us right now. That kind of thing could happen.
Matthew: Almost to a tee, exactly.
Andrew: That’s exactly what would happen. All right. When we talk about chatbots, this is the ideal experience that can happen today. In the future, something else could happen, but today, this is it. I think one of the advantages of this is I might feel a little bit reluctant to talk to another human being and admit that I’m having trouble financially paying off my loans. But talking to a bot, it feels safer. It takes away—the example I keep referring back to is the one from I think the U.S. Army that I heard about on NPR years ago, they had a chatbot for new recruits.
New recruits will not talk to—dudes will not talk to another dude and say, “How many pairs of underwear should I bring on my first day?” because you feel like a sissy for doing it and they feel like a wimp. They’ll go in a chatbot and they’ll ask that kind of a question and they’ll get an answer. That is the kind of experience that I think bots are exceptional at. Questions that come up often that a human being shouldn’t have to repeatedly answer and that people would rather address to a machine than to a human. True?
Matthew: Couldn’t agree more. It really goes back to the whole thing of you briefly talked about this in your intro. Chatbots are a completely new medium right now. It’s a new market. It’s not a mature market, but there are strengths and weaknesses. One thing that a chatbot is extraordinarily well at is allowing people to feel safe. I have nothing to lose by telling a chatbot about my financial situation no matter how bad it is because I know a chatbot isn’t going to judge me.
Andrew: Right.
Matthew: If you can play internet of things that, you’ll see great results, which we’re seeing.
Andrew: Let’s talk about numbers. I’ve kind of trained my audience to care about numbers. So before they had a chatbot, they had the standard landing page. They had ads that were dialed in. They had people who were ready to take on the work that came through that. What was their cost per lead at that point and then what’s the cost per lead now with this chatbot?
Matthew: Yeah. So, for a qualified lead—and just to clarify what a qualified lead is—this is someone who clicked on an ad, got dropped off to their landing website, filled out the form, which is essentially the questions the chatbot was asking you and said, “Call me.” They were paying roughly around $18, which is an industry standard. Arguably, that’s even on the lower end of what some other competitors are paying for this.
Andrew: And now $18 per lead for debt consolidation. What is it now?
Matthew: Now we’re seeing anywhere between $4 and $7 a pop. That is mind-blowing for that industry.
Andrew: Yeah. Part of it is that it’s a better medium for this than a landing page. The other part is it’s novel. I can’t name other debt consolidators who are doing this. I think that’s the advantage of jumping in and doing this today, doing it when they are.
I’m kind of hesitating, because there’s something I want to go back to that we talked about a moment ago, which is this sense of privacy. People do feel more comfortable because they’re talking to bots. They are and, for the most part, human beings aren’t going in and reading it, but I have to open and say to the people listening to me, as a business owner, I see every freaking interaction I have with my bot that happens on Facebook Messenger—and we do have one, I created one for Bot Academy, you can see at BotAcademy.com.
Everything that you type in, including if you make sexual remarks to the bot, I can see it, or if you answer something like, “I have huge debt,” the company that operates it can see it. More than that, there is no anonymity in this as opposed to the old chatbots that happen on the web, the one that I’m talking about with the military. It happened on the web, there’s anonymity. Here, there isn’t anonymity. That’s for good and evil, but it’s just the way it is.
A business owner, the benefit of that is you’re getting a real person, you know who they are. They know my name is Andrew. I didn’t fake it, where I might fake it if I went to a landing page. I might give them a bad email address if I went to a landing page, like Mailinator email address, because I’m not feeling comfortable yet. So a business owner, it gives you real data, real usage. As a user, you should be aware that when you’re talking to a bot, it’s not complete anonymity. If anything, it’s complete whatever the opposite of anonymity is, completely directed at you.
I’m surprised that you got into this business. I’m surprised that you were willing to do this. I think you told our producer one of your last companies was—I don’t want to put words in your mouth.
Matthew: Contextual?
Andrew: You said it was a shit show.
Matthew: Contextual was actually the one that wasn’t as big of a shit show. CrowdHype is the one you’re—yeah.
Andrew: The thing is that you’re going all in on this. You had a bad experience in business. I’m surprised you would go all in. Let me take a moment to talk about one of my sponsors and then we’re going to come back in and ask you why you went all in on this. The sponsor is a company called DesignCrowd. There are times in our business careers when we need small designs made.
For me, for Bot Academy, we teach people how to create chatbots. We certify them. Bigger businesses are now hiring our certified consultants, and they want to show on their website, “I’m a certified Bot Academy graduate. Andrew stands by my work. You can take whatever Andrew’s reputation is with Bot Academy and trust that I’m good based on that.” In fact, ManyChat now is giving our certified graduates a lot of business by referring bigger businesses who want chatbots done.
Well, the thing is that I don’t know how to freaking design that. I don’t want to work even with my designer whose work I like to do it because I want more examples. I want more ideas for whatever this badge is that our certified professionals will get to put on their website. So I went to DesignCrowd, and they had I think it was four questions that I had to answer, like, “What are you looking for? What needs to be a part of this?” A couple of things like that. I filled it out and I went home. For some reason, I always do this on a Friday with DesignCrowd. I came in on a Monday having forgotten the whole thing. I clicked in and I saw a bunch of designs.
I’ll be honest with you. I had like 150 designs. I think 100 of them were shit. The reason is these guys created a brand new logo for me, and I don’t want a brand new logo. Take my freaking logo and just attach it to this certified badge that we’re going to give our certified professionals. So what I did was I went in and started giving everyone feedback. I said, “I don’t want a new logo, use my current logo.” I rated the ones that I liked best so that everyone got to see what my feedback was.
I adjusted my brief to say, “If everyone’s seeing it this way, I can’t blame all these designers who are giving me logo ideas. It’s probably my problem.” So I went in and I adjusted my brief to say I don’t need a new logo. Please use our logo. Here’s a link directly to our logo. Now all these designers are improving their work and I’m getting even better designers and I think we’ve got someone who we’re going to pick from this.
The idea behind DesignCrowd is a crowd of designers submit their designs to my specifications, I give them all feedback, and then the one that I love best is the one that I pay for and get to use. They guarantee that you’re going to get what you want. I think as someone like me who’s not a designer, who doesn’t even have the language to communicate what I want and what I don’t want, their process makes it really easy.
I’ve talked about DesignCrowd for a long time. I’ve used them multiple times. I urge you to go check them out, but don’t go to DesignCrowd.com. Frankly, I won’t get any credit for it, but more importantly for you because you don’t give a rat’s ass about me giving credit, you will get up to 100% off your design projects if you use this special URL that would give me credit. It is DesignCrowd.com/Mixergy.
To give you an example of what that cost me, it’s roughly like $200 to get at this point over 200 different designs submitted to my specifications and adjusted based on my feedback. If you go to that URL, you will also see the work they did for me to give me a new podcast cover art. I think you’ll even be able to click on and see how I asked for it and what I paid for it and so on. So even if you just care about Mixergy and you want to see how I’ve used this, I think it’s an interesting URL to go to. Go check it out at DesignCrowd.com/Mixergy.
Hey, CrowdHype, what was the shit show over there?
Matthew: Yeah. Looking back, you can identify things that weren’t obvious when you were living through it. Just typical early startup stuff, right? I think we got semi-lucky with our first startup in that we took some VC money from a respected VC.
Andrew: You’re talking about Contextual?
Matthew: Contextual, yeah.
Andrew: You know what? Since we’re going back in time, let’s go all the way back in time and bring us to date, and then I want to understand how Black Ops started with getting clients, why it didn’t work out for you and where you lost money, how you found good customers for chatbots. But I think if we’re going to get to know you a little bit, let’s go all the way back to what your mom did. What did your mom do?
Matthew: Yeah. So my mom cleaned hotels for a living. We grew up in a pretty poor environment.
Andrew: Were you embarrassed by that?
Matthew: At the time maybe, but now I wear it like a badge of honor. It really influences a lot of how I look at the world.
Andrew: How?
Matthew: It’s a constant reminder that in what we do as far as being entrepreneurs, you were owed nothing. That is a scary thing, but it’s also a very beautiful thing because it means that you can have anything.
Andrew: What do you mean owed nothing?
Matthew: So you’re not owed customers. You’re not owed revenue. You’re not owed things that typically we see in other parts of the world. Becoming an entrepreneur, there is no formula. Everyone has their own flavor for making it work, and typically what I see and what I see in myself is that the more unique the flavor is, the happier and the faster you grow. That might be a very high-level abstract way of looking at it, but it is a constant reminder. I’m just grateful for where I am today.
Andrew: You used to go with her when she cleaned hotels?
Matthew: Absolutely. We couldn’t afford a babysitter. I remember vividly going with my mom sitting on the end of the bed and for 20 minutes, every 20 minutes she had to clean a new hotel.
Andrew: What would you do while she was cleaning up?
Matthew: Pop on the TV, watch “Ren & Stimpy,” which by the way, no child should be watching “Ren & Stimpy.”
Andrew: No adult should be watching. What the hell is that show? Frankly, while we’re at it, “Spongebob Squarepants,” who’s creating these freaking kid TV shows?
Matthew: Yeah.
Andrew: So you do that, and then one time, you saw that the hotel had these piece of shit computers, as you told our producer. What made them so crappy, and how did you come across them?
Matthew: These were point of sale computers from the ’90s.
Andrew: Oh, I see. I thought it was piece of—
Matthew: Yeah.
Andrew: I’m cursing a whole lot here. I must be under a lot of stress today.
Matthew: It’s cool. I’ll curse with you. So these were piece of shit point of sale—
Andrew: That’s what POS means in this context, point of service. These are something you use at the reception desk.
Matthew: Absolutely. They were throwing them away. My mom saw that. She said, “Hey, I’m going to take one of these home for my son. I think he would enjoy it. She brought it home, 11 years old at the time. I didn’t even know that this thing—I didn’t know how to turn this on. So instead of turning it on, I took it apart, put it back together. I probably did this five or six times just learning how a computer from the inside was structured. I fell in love with hardware. I fell in love with the inside of a computer at a very early age.
Andrew: It demystifies it to be able to do it without having any danger of losing someone’s data on it.
Matthew: Exactly. I really had no idea. This is before I have ever plugged in the little phone cable and got on the internet. This is before I understood the impacts of the internet.
Andrew: To bring it back to chatbots for a moment, when I got curious about it, I went to ChatFuel and they let you make an infinite number of chatbots. I would kind of do the same thing, create one, take it apart, see one of their sample ones, play with it, delete it, attach it to a real account that I created that’s not at all visible to the rest of the world, just explore it. There’s something about that exploration that really teaches you by letting you break stuff.
Matthew: Yeah. I think it’s rooted in the need and the want to be curious. And being curious can apply to a hacker culture. It’s deep inside of our culture, everyone, not just developers or hackers, but even people who do sales. I want them to be curious. I want them to find their own unique way of doing things.
Andrew: Okay. So, fast forwarding a little bit, you worked a company called Killswitch, which has many big clients including Viacom, the parent of MTV and VH1. Somewhere in there, you were messing around with early chatbots. What did early chatbots do, and how did you mess around with them there?
Matthew: Yeah. So when I was at Killswitch, I actually created some of the first early social networking sites. This was back in Myspace days.
Andrew: So they would create a social network for their product?
Matthew: Real World casting website I created, Jackass World online, RuPaul’s Rude Awakening, My Super Sweet 16.
Andrew: Why would they create a whole social network around it as opposed to just a form that people who wanted to be a part of the show would fill out?
Matthew: Because they wanted to gauge how people would actually perform on the show. So, for Real World, they didn’t know that people really wanted to see an 18-year old female in this part of the world put next to a 32-year old someone on this side of the world. At that time, they used it as data leverage to construct who would actually end up making it onto the show. But in that, I learned how to make a social networking site, which was great. I took that all the way to where we are today.
Andrew: So did you have any experience at that point with early chatbots, or am I mistaken?
Matthew: No. So my chatbot kind of education/exposure came a little bit later.
Andrew: So let’s go forward. The first business you had was Contextual. I’ve got a line here that I found on AngelList to describe what it did. You guys at the time said, “Our cloud and mobile-based platform helps companies share important information that’s locked in planning spreadsheets, walled off in expensive enterprise platforms and lost in the black hole of the corporate inbox.
Basically, companies have all this data spread everywhere. Nobody knows what it is, and what they end up doing is asking someone on the team, “Hey, where’s this data? What’s this?” This is the same issue that Slack has noticed, and they’re trying to solve by having on search box for all your company data. It’s still not a solved problem at all by any means. This is the thing that showed you that you—I’d love to do an interview with you just on that, so we don’t go too deep into it today, today I want to focus on chatbots—but I feel like you learned something that you brought to Black Ops today. What was it that you brought with you to your experience today?
Matthew: Yeah. That hits it right on the head. I like to describe Contextual as Slack before Slack. Not that I’m big into comparisons, but at the time, we didn’t realize that you could take something so complex and expose it and share it in a messaging platform. So we were threaded. We took the very old school look of create a space called accounting and attach your AR and PR metrics and pin it on a board and people would manually have to go out and get it.
What we realized towards the end of that startup, a little bit too late as the market kind of showed us, is that everything should be much lighter. Instead of having a threaded conversation, it should just simply be a chat.
Andrew: I thought you had that. You had something that reminded me a little bit of the early Twitter website with the input box at the top, text to people below. What’s the one that was the business version?
Matthew: Yammer.
Andrew: It looked like Yammer in some ways.
Matthew: Almost identical to Yammer. But still, what we realized is that was too heavy for our—
Andrew: Even that was too heavy.
Matthew: Even that was too heavy. That was a key learning towards the end. I’d like to use this analogy. We could have been much more successful had we have sold this thing like a hot rod. Go to Wall Street, find all the hedge fund managers, teams of 20 to 200, have them put it on their Amex discretionary budget and this is the cool thing. Open up the app, one two-line sentence to your team and everyone’s connected rather than the cell of top-down and trying to sell this very complex enterprise system.
Andrew: I see, where someone eat the top of the business has to say we’re buying this. Everyone on the team needs it. Let them just go and buy it for themselves, use it to chat with each other. Frankly, that’s the version that Yammer went for, right?
Matthew: Yeah. They were successful because they had those types of accounts under their belt. What Slack did was the inside out sales model and that worked. That’s the model we should have gone had we have stayed a course.
Andrew: I’ve heard that when people talk about you, this is the big sale of your life. My sense is it wasn’t generational wealth that we got out of this.
Matthew: It’s not. By no means we it fuck you money. That’s okay.
Andrew: What did you exit for?
Matthew: So it’s an undisclosed number that I can’t disclose. Let’s just say that our investors were happy. It didn’t make any founders wealthy. It didn’t make any founders retiring off their island.
Andrew: What did you learn from this then?
Matthew: We learned—
Andrew: You personally, what did you bring forward to your current experience at Black Ops from what happened at Contextual?
Matthew: Well, don’t defer revenue, number one. That was a very, very rooted lesson that we brought over to Black Ops. That’s why we started to generate revenue on day one. Don’t always believe what VCs have to say. I definitely don’t want to pick a founder/VC battle, but I’m much more reserved in this day and age on who, when and why I would ever raise money.
I’m a very firm believer that you shouldn’t raise money to stay alive. I feel like that’s a recipe for disaster, your back is against the wall, you should raise money because you found something that fundamentally works and you need to get there faster. That’s an amazing position to be in. I think you find partners that are much more aligned with that vision.
Andrew: The next thing you did was CrowdHype. Again, I’ve got a line that I think you guys put online somewhere about it. It’s connect and engage with fans and friends around live events. A live event could be a concert I go to, where I can talk to other people there, but it can also be iOS 25 coming out. I want to see what people are thinking about it. I want to talk to people about it and this is what I would have used your app for, right?
Matthew: Correct, Twitter moments.
Andrew: I see. When you say it was a shit show—again, I’d do a whole interview just on that—what do you mean? What happened?
Matthew: Yeah. So, I think it means a lot of things, but I think most importantly, it means that we had some very big partners where the vested interest just wasn’t there. We had some partners that the intellectual property that we had licensed to be able to do this which got us access to media and tickets—by the way, anyone in the ticketing space knows that this is a mafia, not a legal mafia, but a very close held, single-digit number mafia.
Andrew: Where a small group of companies own the vast majority of tickets because they own the ticketing and own the locations.
Matthew: Absolutely. We got to know these people. Actually, the company that held the intellectual property invested in our company, which ultimately killed us when they got killed. It was almost like a house of cards in that we had a great implementation on a great idea. We had extraordinary user value that was being unlocked inside of the app, but because of the order of operation and because our app depended on the intellectual property of a company who was funded by bigger companies, as it started collapsing, we were left there on the bottom.
Andrew: So what’s one thing you took from that and brought to Black Ops?
Matthew: Yeah. Again, do not defer revenue and stand tall and stand solo. I think a lot of startups get this mentality that they need to have partners. They need to have big partners on their deck. They need to have big advisors on their deck. That’s simply not the case. If you have something that is truly valuable, it doesn’t matter who your partner is. It just matters how much energy you have to go sell it. All of that was almost a fiery storm that started Black Ops.
Andrew: When you started Black Ops, you were thinking, “I’m going to create a digital agency,” right?
Matthew: Correct.
Andrew: Why a digital agency? We’re living in a world where SaaS scales so well. You create software as a service, you get customers. It’s painful for the first year or two, but if it’s a good product, people will eventually pay you enough that you can fund the business, you can keep momentum going. But when it comes to digital agencies you have to do the work. You’re selling hours, your hours, other people’s hours, it doesn’t scale. One customer leaves you and you’re screwed. Why did you decide to go into this?
Matthew: Two reasons. The first reason was over the last four years of doing startups, we met amazing people. We met CMOs, CEOs, board members for big companies and created really rich connections with them, but never found the right opportunity to work with them. So creating a digital agency provided a house for me to go out and re-contact all the people we’d met in the last four years and say, “Hey, we can work together now. What do you need?” which turned out to be a very great early sales strategy for us.
Two, we knew we could do it differently. All of us were startup founders. All of us were hackers. We knew the digital agency market was timed for almost a hybrid model, which now we do to almost a gross level, but even then when we were making mobile apps and websites and APIs, we created frameworks that we reused. So it was less about selling hours, and it was more about selling a reusable product that we saw a hole in the market.
Andrew: I see. You’re right. There are a few companies that do that. Roger Dickey, a past interviewee here, started a company, I forget the name of the company, but essentially what he’s doing is saying well, everyone needs a login screen, but if you need a login screen, we already built it, here’s how much it’s going to cost you to have that incorporated into your app. If you need a list of customers or list of users, we built that already. We can add it in for this price. That’s what you were thinking, modular development, keep reusing the pieces. Is that the kind of thing?
Matthew: Absolutely. The most important thing there is that login system is probably the best login system you’ll get because it’s been battle tested, right? It’s been redeveloped and improved and incremented. So one of the things that a lot of people don’t understand with a digital agency until later in their years is what you charge is value perceived. You don’t have to charge per hour. If I make a login screen and it takes me 20 minutes, it doesn’t change how much value that is to you, my customer.
Andrew: Right. Why should I, as a client, have to pay for every minute that goes into it? I want the result. I want the login screen.
Matthew: Exactly.
Andrew: So I see the vision here. I’m looking at the Internet Archive, and I see how your business transformed from being not super-specific to suddenly there’s a big link on your site saying, “Here’s what Facebook chatbots are.” So at some point you found the religion of chatbots. It seems to have happened around this fashion brand that you worked with, a company from New Zealand. Can you say who they were?
Matthew: Yeah. They were called Rodd & Gunn.
Andrew: What is it, Rot and Gun?
Matthew: Rodd & Gunn. They’re headquartered out of New Zealand, huge in Australia, New Zealand. I believe they’re an $80 million a year fashion company. They were coming into the States last year. I think they delayed that to this year a little bit. But they’re opening up 80 stores in the U.S. We were building their mobile app, their backend system, and we were helping with the transaction to their new point of sale system.
Andrew: I see their site. I thought it was going to be a women’s brand. It’s for dudes.
Matthew: Yeah. They opened up their first store in Brooklyn. I believe it’s in Dumbo, actually, right up the street from me. They’re compared to Ralph Lauren.
Andrew: Yeah. I can see that, maybe a little bit more conservative, but the kind of thing that you’d envision yourself taking to a country club, but you can also just go into a local bar wearing and feel comfortable.
Matthew: Yeah.
Andrew: So you got to work with them and they said, “Hey, there’s something you need to go explore.” This was in April 2016. What did they say you needed to explore?
Matthew: To set some context here, this was actually a week before the F8 event last year. So this was before Facebook even publicly mentioned anything about bots. They had called us and said, “Hey, Matthew, we just got off the phone with Facebook. They told us about this super cool thing called bots. Can you do us a favor? Look into it?” “Of course, absolutely I’ll look into it.” So we got access to some API docs, which I cross-referenced with one of my developers. We thought it was very interesting. You didn’t quite see the full the vision. We thought it was very interesting from a distribution method, and we kind of shelved it there.
Andrew: This was the chatbot platform that Mark Zuckerberg ended up announcing 2016 at F8.
Matthew: Exactly.
Andrew: Let me take a pause right here and come back into the store and pick up from there. The pause is to talk about my sponsor, Toptal. Before the interview started, I asked you if you have any issues with Toptal, do you invest in a competitor, which sometimes happens so I wouldn’t bring it up. You said, “No, actually, I’ve used them.” So, I wondered how did you use Toptal, the company that anyone can go and hire top developers from?
Matthew: Yeah. So, as a digital agency, sometimes we have influxes of work and just so happened to have an influx of work a couple months ago, needed a Ruby on Rails developer, went on Toptal because I heard great things. It’s really cool because what they do, you have an account person but that account person is super technical. It’s almost like a CTO. You talk with that person. He really understands from a technical standpoint what you’re looking for and then he handpicks three people for you to talk to and all the people he handpicked our CTO paired the interviews with us.
He didn’t know which person to pick, so we just picked the person we thought was the coolest and it worked out really, really well. We were a little worried about hiring someone from the opposite side of the world, how would the time go, the communication, time difference, communication difference, but we experienced none of those problems.
Andrew: Yeah. I’m surprised at how well they do that. I just talked to someone from their team who happened to be from San Francisco. I said, “Are you coming here for a business meeting?” He said, “No, I just travel the world.” He says, “I was in Iceland before.” “Isn’t that tough?” “No. We have specific hours that we work. Everyone knows when they can count on us and then the rest of the time I’m exploring a new country.”
So one of the things that I’ve understood from digital agencies is they often need the developers to talk directly to their clients and so did you have that, where the developer you hired needed to—no, it was just with you guys. You’d give them the work, the developer would do it, give it to you and then you’d interact with your clients.
Matthew: Absolutely. Yeah. We run a pretty tight ship in that we’re boutique for a reason. When we do a project, almost the entire team is focused on that project. I have pretty much the back and forth with the clients being a salesperson, a technical person, I also can bring that back to my team as well.
Andrew: They’re not the cheapest though.
Matthew: They’re not. That’s actually a selling point to me.
Andrew: Why?
Matthew: I want to know that one, the person I’m hiring is making a living wage, but two, you get what you pay for. It’s no different for any product you buy or any person you choose to hire. I believe in paying well because I want to have a service done well.
Andrew: If you guys are out there and you want to sign up for Toptal, it’s created by two Mixergy fans. One of them did an interview here. The other I’m reluctant to ask to ask to do an interview because I think he’s so secretive about everything. He’d be upset by any question I’d ask. These guys take their business way seriously. I want to get drunk with one of them. I’m not even a get drunk type of person. I would drink to enjoy the drink. I would like to see what they’re like when they relax because these people are freaking intense, the founders of Toptal.
Anyway, they’re Mixergy fans. They’ve created a URL where you can sign up and get this kind of service, get these great developers. In addition to that, you’re going to get 80 hours of Toptal developer credit when you pay for your first 80 hours and that’s in addition to a no risk trial period of up to two weeks. You’ve heard many Mixergy interviewees here tell you they’ve signed up and worked with Toptal or they’ve heard me talk about Toptal and they took down the URL and they went and signed up. I urge you to go do it too. It’s top as in top of your head, tal as in talent, Toptal.com/Mixergy.
I don’t think I’ll be able to hang out with or drink with them for the next ten years. These people are so freaking insane—so addicted, so focused on what they’re doing. They raised money from Mark Andreessen. I asked this guy, “So are they going to do another round? Things are going well?” He goes, “They’re doing fine.” I don’t know that I should be saying that out loud. They are just—you know what? I love people like that. Do you feel like you’re like that, where you have this passion for your business, for Black Ops that it’s all encompassing?
Matthew: Yes and no. The no part is something I’m doing more recently now. Had you have met me a couple years ago, you probably had noticed that I work at least 90 hours a week. I would be 100% all in and it was that way because I wanted it to be that way. I found something I was super passionate about. I would go head’s in, complete focus, no disruptions in my life, no extracurriculars. What I realized is that there is no way for one human to be working that much and have the highest quality of output. You can work 90 hours a week, but how many of those hours are sloppy work? How many of those hours are spent doing things that could probably be done in half the time if you were sitting down and focused?
So now I’m about to turn 30. That’s a different chapter of my life. I sit down and I almost time box myself. I wake up super early still, but when the time comes to work, I fucking work. When the time comes to disconnect, I disconnect. That is part of my whole system. It allows me to recharge, allows me to have those high-level thoughts so when I come back tomorrow, I’m refreshed and ready to go.
Andrew: I’m experimenting with around the house not having my iPhone with me because I’ll get sucked into junk and using my Apple Watch so if I need to—stuff will come up, like I have to go and do this. Instead of doing it right away, I’ll just into the Apple Watch say, “Siri, remind me to do this tomorrow. Remind me we need this and buy it from Amazon tomorrow.” And it really helps me not get sucked into junk work that really is not very efficient at 9:00 on my phone.
Okay. You discovered this chatbot thing. You see it before it’s launched. Mark Zuckerberg launches it at F8. You do what next?
Matthew: We go all-in.
Andrew: Meaning no more of this other thing you built?
Matthew: Not yet. I’ll get there. We go all-in in that we created one of the first chatbots that was on the platform. It was an unofficial UFC chatbot. We did this because we were obsessed with UFC internally, but more importantly, we did this because we found an unpublished API. What had happened—this was before cards. This was before any of that stuff. So the experience was hey, type something and we’ll return something back. It was terrible.
But what we learned is that people loved interacting with this thing. It could have been because the demographic was I’m a UFC fan. I want to poke around with something. But we actually ran a little experiment before Facebook ads were able to point to chatbots. We went to Reddit, bought out the entire subreddit for a weekend of UFC, it feels like UFC 150-something. We ended up getting users at $0.03 from Reddit who manually went from the web to the Messenger and interacted with the bot. That’s how engaged people were back then because no one knew what a bot was.
Andrew: What do you mean? They’d have to go to Facebook Messenger, search for Unofficial UFC Chatbot.
Matthew: Yeah.
Andrew: I’m on the chatbot right now. Then they’d start talking to the chatbot.
Matthew: Yeah. They would actually go to Reddit. We bought out the subreddit. It was $153. That’s it. I think our ad was, “Come talk trash to the UFC chatbot.” People clicked on that, went to our landing page, from our landing page went to the chatbot because there was no direct URLs at that point.
Andrew: Oh, I see.
Matthew: It was so new that people had to go through all these hurdles but people still went through those hurdles. No one went there for the actual thing we were developing, which was go to talk to the chatbot, say Conor McGregor, get the stats on him, see when the next fight is coming up. Everyone went there to see how a chatbot worked. People talked trash to it. People said, “What do you do? What is this? Are you a human? Are you real?” That was so promising to see, but it was so painful to see the experience wasn’t there yet.
So that’s how we started. We did two or three of those types of bots, where it was strict conversation, say hi, say this keyword, return something back. The experience wasn’t there, but we were smart enough to know because we had gone through this evolution so many times—web to mobile, mobile to now conversation that we knew this would take time. So we went heads down. We didn’t release any bots for probably about eight months, but in those eight months, I actually went out to New York City and I started talking to people. I talked so very long because I—
Andrew: You told our producer you went to was it Madison Avenue talking to Fortune 500 brands about bots.
Matthew: Not just Fortune 500s but also the biggest agencies.
Andrew: How did you get into Fortune 500s to get the right person to tell him about chatbots, agencies. How’d you get in?
Matthew: I cold emailed them.
Andrew: You cold emailed and said what?
Matthew: So, I had a process. I actually don’t mind sharing that. So, I created scripts and the scripts were connected to two things. One was an enrichment service, Clearbit, awesome service.
Andrew: Yes.
Matthew: The other was HubSpot, which would pop out sequences, which was essentially their version of drip campaigns. I would target CMOs—I was smart enough to know that no one knows what a chatbot is. No one knows they need a chatbot, but everyone knows that they needed improvement on their ROI. I knew where Facebook was going with this. I knew that this would be something that was able to expand the advertising ecosystem.
I knew based off of what we were seeing that this would be a better engaging tool, that you’d be able to do clever things with marketing and targeting. So I targeted the CMO and I said, “Look, we have a way that is converting customers cheaper and engaging them at higher rates. Are you interested? I can come over and show you more.” I did that so many times. I literally cannot count the amount of times everyone would say, “That’s awesome. Come over.”
And then I would tell a little bit about chatbots and there would be this blank face. Every single time I got a little bit better understanding where chatbots actually fit in an organization. Some people could say sell them to the CTO, some could say sell them to the social media team. But where I found ultimately is it fit in the marketing budget and you had to target CMOs who were already spending significant amounts of money buying their customers online and if they were doing it socially on Facebook, which now we have a script to determine those people, even better because all you can say is today, your destination is here. Tomorrow it’s going to be here. Here are some actual results you can expect when you do that.
Andrew: You didn’t have any results though. It was just here’s what I think you could expect if we do this.
Matthew: Yeah. At the time, we didn’t have any results specifically around what—we call them conversation funnels, which are conversion funnels, what that would actually do from an ROI perspective, but we knew that it would increase engagement. That we had saw at an unbelievable way from our hacking around.
Andrew: Why would they see you? You’re just a guy. They don’t know if you’re a college kid with an idea or someone who’s at a competitor or any of it. Why would they see you?
Matthew: I think that’s a great question. I keep coming back to the same answer. I think the answer is that a lot of Fortune 500s right now are scared. They’re scared because they see their user base adopting things at an unbelievable rate and they see the pace at which they’re able to actually produce. It’s not in sync anymore. A customer’s adoption to something like Facebook Messenger whereas, the brand is just getting their heads wrapped around marketing advertising.
So I think a lot of it was just purpose curiosity, what does this guy have to say. I’ve kind of heard those words before, but I’m still not sure what it is and this could be used as a way to understand it, whether we hire this firm or not, at least us as a team understands it. One of the things that was consistent across all these meetings is it wasn’t just, “Hey, I sent an email to the CMO and I had a one on one meeting with the CMO.” It was, I sent an email to the CMO and the CMO invites their entire team.
Andrew: Just so they could see what is this new thing. You didn’t even use the word chatbot when you messaged them?
Matthew: I did.
Andrew: You would. You’d say, “Here’s a way to increase engagement,” and what was the other one?
Matthew: Have an ROI increase on your advertising spend.
Andrew: I see. And chatbots are the future, can I come in and talk to you about it. The fact that you’d come in is a big thing as opposed to getting on a call.
Matthew: Absolutely. I’m a face to face kind of guy. I hate telephone calls, a place in time for them, obviously. But I’d much rather be in front of you, shut my laptop and just talk.
Andrew: I’m noticing the phone calls are incredibly inefficient for business conversations because people will do other things if you talk to them on the phone. So more and more people will Facetime me or Zoom, Skype less so and video is more work, but it’s more worth it. Okay. So, you do all this stuff. Let me understand this whole Clearbit thing. Clearbit just takes the email addressees you have and gives you more data on them.
One way that I’ve seen people use it was they would take their email list, run it through Clearbit, find out who their big clients are and then message those people, the people who work at big companies and say, “Can we get on a call with you? We have this added service that you may not know about and we can help you with it.” Then they upsell them. How did you use it? How did you get the email addresses that you were reaching?
Matthew: I actually started with their prospecting tool, which was essentially I would narrow target—say I was going after agencies. I would so agencies of $20 million in up headquartered in New York City. I would get the list there and then out of each of those agencies, I would say who’s the CMO, if I couldn’t get the CMO, then who is the head of digital marketing.
If I couldn’t get that person, then I’d go to someone like a senior social person. I had this system where I would just constantly prioritize and constantly just write these scripts to pull these people out. And then every email I sent, while it did go into a sequence, I actually had it drop into my drafts so I could hand read over it and then I would do some research, like go to their website and put something personal in there.
Andrew: I see.
Matthew: Go try to find the person on Twitter and see that they were at an event that I was out and call them out like that because that’s the stuff that actually works, right?
Andrew: Yeah. I find there’s a lot of software right now that will send the first message out and the second one will come back afterwards and say, “You may not have noticed this. I’m responding to get it to the top of your inbox.” It’s easy to do that. It’s so generic. I just immediately respond stop.
Matthew: Exactly. It has a negative effect.
Andrew: That’s how you started to get customers. You used the chatbot as—
Matthew: Well, I didn’t start getting customers. I just started getting our name on the map. This was an eight-month period. This wasn’t a sell to me. This was an educational period to me. We knew the market wasn’t ready to build very compelling bots, but we knew it would eventually get there. My whole play was if I start now and all these larger agencies, all these larger brands know who Matthew Black and Black Ops is. When the time is right, they’re going to call me. I just constantly have that in mind. It sucked because we deferred a lot of revenue. We almost went bankrupt. We lost some key players. But at the end, it was the exact right thing to do because that’s happening now.
Andrew: You told our producer, “We went broke.” You ran out of money and the reason is you stopped looking for clients for the previous business we described where you’re going to build apps for people using the same code over and over and instead you just went out there and taught people what these bots were on the hope that when they’re ready that they will buy this and on the hope that Facebook won’t kill the platform and they will be ready.
As a result, we talked about one of your clients, you got the Debt Like WTF. Who were some of the other companies? I’ve got them here on my screen, but I’ll you tell me which ones you feel comfortable talking about.
Matthew: Yeah. We’re working with BarkBox, which is awesome.
Andrew: BarkBox?
Matthew: Yeah, they’re like a subscription box for your dog. Every month a box of toys comes up. They contacted us after we got put on a 2017 it’s called Rock Star Agencies to Watch list, didn’t think anything about it at the time, but it coined us as the chatbot agency. There was no other chatbot agency. We didn’t realize how powerful it was to start getting coined that way.
Anyways, BarkBox contacted us and they had a cute little idea of saying, “We want to build this cute little SMS bot where you pretend to be your dog and it walks you through a little game and at the end, it gives you kind of like an Instagram filter and people will share it to their Instagram.”
Fast forward, it was the most engaging thing BarkBox had ever done. Seven days in, over 100,000 messages sent out and received. The response was so great internally that we’re looking at laying out a 2018 vision, which includes a monthly recurring game, a way for SMS to actually have an upsell utility. So you’d customize your bots and a few other cool treats that I can’t talk about.
Andrew: When you say it’s the highest engaging thing, is that something that they could compare? It seems like an apples to oranges comparison. You can’t say how are people engaging with this versus Twitter. It’s a different kind of engagement. I’d be less likely to want to engage with something on Twitter and I shouldn’t even though I might care about it than I would in something that comes into Facebook Messenger that I have to respond to.
Matthew: Absolutely. I think when they say the highest engaging initiative they have, they’re comparing it to things like Instagram likes, things like Twitter likes. While I do agree with you, it’s not an apples to apples comparison, BarkBox is an interesting company in that the most important thing to them is customer engagement. That’s how they gauge how well they’re doing as a company. They’re not focused like other companies are on just turning out customers. They’re focused on holding on to customers.
So that was—I think the story there is one, they got into this with some hesitation. Some people were scratching their heads, “What is this? Is this going to even be an important thing to our company?” Fast forward to today and it’s one of the most important initiatives in the company.
Andrew: Okay. So BarkBox, what else? Who are a couple of other customers that are coming on board?
Matthew: We’re launching an awesome bot later this week for Haven Life. Haven Life is owned by MassMutual, one of the biggest insurance companies out there.
Andrew: What do they want with chatbots?
Matthew: That’s when you know things are working, the education I was talking about started to pay off. So Haven Life is kind of coined as the [inaudible 01:02:43] for term life insurance. They had saw what we do with the debt company and decided that we’re already buying the majority of our customer through Facebook. Let’s build a chatbot to qualify and give someone a quote directly in line.
We built an experience that actually educates someone on what’s the difference between a $500,000 policy and a $1 million policy and what’s the difference between a 15-year term and a 30-year term because, ultimately, no one really understands the difference, but through a conversational interface, we’re able to easily articulate those things and we integrate with their backend API to display a quote directly in line and they’re able to continue that quote.
Andrew: You’re quoting all this stuff yourselves. You used to use Chatfuel, you’re not using Chatfuel for all these. One last customer you’ve got coming up, who can you talk about?
Matthew: Humana. This one is my favorite to talk about because of one reason, the demographic. The chatbot we’re building is for the open enrollment for Medicare in November. This is like a Super Bowl for them. At any given time, they have up to 1,500 concurrent phone calls of people trying to enroll for Medicare.
Andrew: Just during this open enrollment period.
Matthew: Yeah. That’s insane. But the target demographic is 65 and up and no one is focusing on that age group although it is the fastest growing age group on Facebook. The initiative there is one, will a 65+ person actually interact with a chatbot? We had to build an entirely different strategy, one that’s more instructional than you would with a younger demographic who’s more used to talking to newer forms of tech. I think it’s going to perform really, really well and ultimately it’s to educate you on Humana’s version of Medicare and then schedule a call and apply for Medicare.
Andrew: All right. If anyone wants to check you out, I’ve got the list of the hot ten agencies you must know in 2017. There you are. You’re listed as the chatbot and AI geeks. I don’t think anyone else on this list has chatbot or AI on their list. So I can see why you’d customers from it. The URL for anyone who wants to skip the list and go check out your company is BlackOps.nyc because you’re in New York City.
Matthew: That’s correct.
Andrew: BlackOps.nyc. Guys, if you’re into this at all, I feel like this is a transformational period. There were these huge companies being built when we went to mobile, huge companies being built when we went to SaaS. This is it right now. This is the time. I’m excited to hear how you did it. I’m thankful to you and to my two sponsors, DesignCrowd.com/Mixergy if you’re looking for a bunch of designs from a crowd of people who are really good. And if you need a developer go check out Toptal.com/Mixergy. We also have a chatbot. Go check ours out BotAcademy.com. You can see what our chatbot is like. I went to your site. I don’t see that you guys have a chatbot.
Matthew: It’s on purpose.
Andrew: Why not?
Matthew: So there’s strengths and weaknesses. We do not believe that a chatbot is good for selling B2B.
Andrew: Interesting. Okay. So what you have instead is—
Matthew: All our chatbots are for B2C customers.
Andrew: So you have a human there. Her name is Hillary right now. I see that she and I have been exchanging messages because I wanted to see is she a bot. She’s a real person. Anyone who wants to go see the real business, check them out at BlackOps.nyc. Matthew, thanks so much for doing this.
Matthew: Thank you, Andrew.
Andrew: Thanks. Bye, everyone.
